The purpose of anything data driven is to provide evidence that helps with decision making.
Cole was a low risk/low cost stop gap that's all.
I'd say they got most of the decisions right for players.
As far as Mowbray is concerned, Nestor talked about "uplift", I assume, by that, he meant the squad was in a better place when he left a club than it was when he joined. Depending on what criteria & units of measure they used, it's quite likely that Mowbray came third of the coaches that could have been available.
It's pretty clear to me, from the section on game model, that Nestor & the coaching team have had conversations on how to get through the low block, so Nestor sees himself as a mentor. That's a good way to move forward IMO.
My educated guess (in that model uplift is something used in my industry) is that they created a model to predict performance of teams on paper, and then looked at actual performance, and the gap between modelled performance and actual is your "uplift".
Without making a load of assumptions it's hard to guess what the mod had in it, but essentially it's trying to say "This season should have finished with 55 points but got 65 points, so the manager has a +10 uplift on performance if all else is equal".
My main issue with that is depending on what the model is trained on, you don't know if it was the manager who caused the uplift, or maybe it was luck in the way the fixtures fell, or something else you haven't considered.
Fans (and football professionals) like to underestimate the role of luck and variance in football. If you have an 80% chance of winning a game, then losing shouldn't come as a surprise because that's still a 1 in 5 chance! (Ignore draws just for the sake of argument).
There are so many factors that can't be considered - refereeing for example - and there are things that don't tend to enter the conversation, like when you play a team - are they on a poor run of form or having a new manager bounce? What injuries do they have?
Attributing all uplift to a manager is a dangerous thing in my opinion, unless they have research on how much impact a manager can have, and how long they need to have that impact